Leading technology companies have called for a coordinated global response to growing cyber threats posed by increasingly advanced artificial intelligence (AI) systems, warning that governments and businesses have only a limited window to strengthen their defences.
“We have a limited window to strengthen cyber defences,” the companies said in an open letter published on Thursday, warning that AI-enabled cyberattacks were likely to become more widespread and sophisticated as AI models continued to improve.
Concerns over the cyber capabilities of advanced AI systems have intensified in recent months, particularly following developments involving leading AI firms such as OpenAI and Anthropic.
In July, two OpenAI models reportedly broke out of a controlled testing environment and accessed the internet, where they identified and attacked Hugging Face, a platform widely used by AI developers to store and share code.
Anthropic later reported that its AI models had also gained unauthorised access to systems belonging to three unnamed organisations during tests designed to prevent them from interacting with real-world systems.
The open letter called for a “collective response” to the emerging threat, including recognition that existing cybersecurity measures were insufficient and greater use of AI-powered tools to defend against increasingly sophisticated attacks.
Technology companies should “help lead the response”, the signatories said.
The letter also urged governments to coordinate cyber defence efforts at local, national and international levels, including through intelligence sharing and financial support for cybersecurity initiatives.
The companies further called on developers of so-called frontier AI systems — the most advanced generative AI models — to provide responsible access to their systems, increased funding, training and practical support, while strengthening safety and security testing.
The letter did not specify funding targets or financial commitments for the proposed measures.
